Tune Harness
Use when harness output has plateaued, the evaluator's grading disagrees with a human spot-check, the same criterion keeps failing across runs, or quality is drifting between iterations. Triggers on "/tune-harness", "tune the harness", "the evaluator is wrong", "the generator keeps missing X", "improve the harness". This is the trace-reading loop — the primary engineering work after instantiation.
Works with: Claude Code (native) · Cursor, Codex CLI (manual)
native: this artifact type is that client's own format
Install (Claude Code):
cp -r tune-harness ~/.claude/skills/- Adoption: 1 repos
- Upkeep: 54.0
- Freshness: 78.0
- Evidence coverage: 84% of the inputs this score can use — the rest are unknown, and the score is discounted for it
- Health: active
- Contributors: 1
- License: MIT
Security audit
Not scanned yet. We audit npm-published capabilities for known advisories, install-time scripts and permission surface; this one has no npm package we can resolve, or has not reached the queue.
source ↗ · skill:kamilseghrouchni/tune-harness
Already running this? npx tashan-cli doctor checks your whole config against the Index — how it works ›
Measured 2026-08-09 · scorer s5 · how · something wrong here?